The Challenge of Value-Based Care in Oncology: Improving Clinical Outcomes

June 2011, Vol 2, No 3 - VBCC Perspectives


The focus on value in patient care has taken a prominent place in any discussion related to improving care delivery, but no agreement exists on what that “value” is. In this interview, Value-Based Cancer Care asked Dr Mitchell to elaborate on the way his plan, SelectHealth, approaches this critical issue in cancer care and the potential implications to payers and providers.

Q: From your payer’s vantage point, what does value-based oncology care mean to you?

Dr Mitchell: When discussing valuebased care in oncology, the first question we must ask is—how do we define “value”? We can discuss value-based care for any disease, be it diabetes and the use of a branded medication in the pharmacy benefit, or oncology, in which the care model would vary all the way to the recently publicized accountable care organization that would involve a bundled or a capitated rate for cancer treatments. Once we agree on what we mean by value-based care, we also have to define the challenges—be they logistics or clinical in nature.

Q: Can you describe some of these challenges?

Dr Mitchell: The main challenge of value-based cancer care for payers and for providers is the application of qualitymeasures. There is a big push today to use pathways and guidelines in the entire oncology arena, from the National Comprehensive Cancer Network all the way down to the local, facility-specific doctor who is following pathways. Putting these pathways into place is a challenge: some organizations measure their pathway compliance and some do not, but very few centers measure the clinical efficacy of the pathway they use.

I have yet to see a study that shows improved clinical outcomes based on pathway use. There are some studies, or white papers, on reduced costs with the use of pathways, but none shows improved clinical outcomes.

US Oncology is likely the biggest model we currently have in terms of published results on pathway use: they measured pathway performance of their doctors within their facilities, showing that the clinical outcomes were similar to not following pathways, but patients who continued to be treated by a pathway had reduced costs, on a per-patient per-month or per-year basis. However, we have yet to see reports that show better clinical outcomes.

This means that the true challenge is the measurement itself—how are outcomes measured? What are the specific components that should go into the clinical or the quality measures? This is not easy to define.

Q: How does SelectHealth address the need to improve clinical measures in oncology?

Dr Mitchell: We have recently embarked on a pilot program to address this very issue of howtomeasure clinical outcomes based on pathway use. Being a regional plan gives Select Health the opportunity to build relationships with local oncologists, with their key opinion leaders or their organization heads. The first step toward improving clinical outcomes is building and maintaining relationships with oncologists, relationships that are based on trust and that allow us to discuss clinical outcomes—what the clinical offices are currently measuring— and then enable us to build objectives that coincide with the clinicians’ and payers’ goals, to align our goals.

This year we are implementing goals around 4 areas in cancer care, trying tomeasure those areas based on performance of pathway use. These 4 measure areas are:

  1. Adjunct therapy for colorectal cancer, and the 3 following areas in supportive care
  2. Erythropoiesis-stimulating agents (ESAs)—red blood cell stimulators
  3. Colony-stimulating factor (CSF) drugs—white blood cell stimulators
  4. Antiemetics.

But that necessitates building a way to measure these, which is not an easy task. To do this we are using both pharmacy and medical claims: based on these claims we are trying to assess whether a treatment fits the pathway criteria. This is often complicated because of laboratory values, staging, diagnosis, previous therapy and concurrent therapy, as well as other items reflected in the claims.

One example is a patient who received a growth white cell stimulator (often used in patients with cancer to fight infection), but based on the patient’s white cell count, that treatment may not seem appropriate. However, the patient may have a reason that would qualify him/her for that treatment, yet this information is not going to be found within the claims. We need measures to verify that what the clinics are doing is appropriate, which is what Select- Health is doing with our pilot program. Figuring out the measuring components is one of the biggest challenges.

Q: Is this pilot program unique to SelectHealth?

Dr Mitchell: Several other plans are working on similar initiatives. One of the most publicized is the Michigan Blue Cross Blue Shield (Michigan Blues) health plan, which is using the P4 Healthcare Oncology tool. The P4 Healthcare is a company that puts together pathways for providers and payers (the 4 “Ps” stand for physicians, patients, pharmaceutical companies, and payers). Payers that are using the P4 tool are reimbursing based on providers’ compliance with the P4 pathways, which are agreed on by academic centers and community oncologist groups; Michigan Blues have agreed on their payment plan in oncology based on those P4 rates.

This regional payer trend is now spreading to national payers that are also buying oncology pathways to structure their reimbursement. According to a press release issued in March 2011 in Specialty Pharmacy News (vol 8, no 3; http://aislicenses. com/NSPN/fetchPdf.php?folder= SPN_pdf&name=spn0311), Aetna is the first national plan to sign a deal with P4 Healthcare for cancer management.

This recent move by a large national payer such as Aetna signals a new trend of national plans signing up with the main pathway providers (P4 Healthcare or US Oncology) to structure their reimbursement plans for cancer care. This trend is likely to continue and grow. This is one of the reasons that the large purchasing groups, such as AmerisourceBergen or McKesson have purchased these pathway-driven companies.

Q: Does this growing trend suggest that pathway use is the road to achieving value in cancer care?

Dr Mitchell: This remains to be seen. Using oncology pathways appears to help reduce the costs of cancer care but to date it has not been shown as a way to achieve value, if we accept the definition of value as including enhanced clinical outcomes.

Indeed, it is difficult to define valuebased care in any area, and particularly in cancer care. How do you define the value of a life? It is also hard to measure performance clinic to clinic, cancer site to cancer site, or physician to physician (eg, some physicians get more difficult patients). It is truly hard to define value in cancer.

A true model of accountable care, however, may be the direction toward value-based cancer care—an accountable care model where providers are paid a certain amount to take care of a patient with a certain cancer stage and a certain type of cancer, and the decisions are based on the patient’s true needs; this may be, for example, not ordering every test that may or may not help, and only doing things that are directed to care, including the use of cost-effective therapy.

Q: What are some of the quality and cost implications for payers?

Dr Mitchell: The first concern for payers in cancer care is—how many lines of therapy should a health plan cover? There is a point of diminishing returns. If you try the first line of therapy for a specific cancer, you get the best outcome. If you have to try a second line, you’ll get a lesser outcome. If you try a third line of therapy, you’ll get a lesser outcome, and so forth. The point of diminishing returns is after the second line of therapy.

The second issue is, if drug A and drug B have very similar outcomes, but drug A is much less expensive, then the ability to use drug A for the majority of the time and cover A for the majority of the time is a big issue. From a payer’s standpoint, it may have to be done by adjusting the reimbursement away from the way reimbursement is traditionally done.

For example, using an average selling price (ASP)-type reimbursement, which is a percentage-plus reimbursement, where drug A costs $100 and drug B is $1000, if you reimburse ASP plus 20% for each, the physician is likely going to use drug B, to get the 20% of $1000 instead of 20% of $100. If we can find a few of these “big therapies” with similar clinical outcomes, then we will reimburse much higher for a generic drug and still save costs.

Finally, no payer wants to end up in the news for saying they denied care because they did not value 4months of life or 2 weeks of life. In pancreatic cancer, there is a drug approved based on 2 weeks of survival benefit. Is this value? In cancer, 4 months is actually a fairly decent amount of time. These are major challenges for payers.